213 Miles From Shea

Over the last couple of weeks we have reviewed the 2015 projections for the Mets for accuracy based on the actual results of the 2015 season. Today, we will start to go through the 2016 Mets and present projections from up to 9 different sources and taking an average. The schedule:

Today – Asdrubal Cabrera
3/5 – Yoenis Cespedes
3/6 – Michael Conforto
3/7 – Travis d’Arnaud
3/8 – Alejandro de Aza
3/9 – Lucas Duda
3/10 – Wilmer Flores
3/11 – Curtis Granderson
3/12 – Dilson Herrera
3/13 – Juan Lagares
3/14 – Kevin Plawecki
3/15 – Ruben Tejada
3/16 – Neil Walker
3/17 – David Wright
3/18 – Bartolo Colon
3/19 – Jacob deGrom
3/20 – Jeurys Familia
3/21 – Sean Gilmartin
3/22 – Erik Goeddel
3/23 – Matt Harvey
3/24 – Steven Matz
3/25 – Hansel Robles
3/26 – Noah Syndergaard
3/27 – Zack Wheeler
3/28 – Addison Reed
3/29 – Antonio Bastardo
3/30 – Logan Verrett
3/31 – Dario Alvarez

This is my third year of doing this, and I’ve made some improvements:
– I’m using an HTML editor so the stats come up as a table and are easier to scan
– This is the most projection sources I’ve used, up by 4-5 from last year (depending on the player)

Bartolo Colon is a hero on this staff of aces. His arm allowed him to absorb innings (he’s below 200 for the regular season only due to the Mets going with a 6 man rotation at different points for the season) and he can do whatever the Mets need him to do. Could the computers predict how he would perform in 2015? Let’s take a look:

PECOTA (BP): 10-10, 28 G, 175 IP, 123 K, 29 BB, 1.16 WHIP, 3.54 ERA
PECOTA (MLB.com): 9-9, 144 IP, 100 K, 41 BB, 1.16 WHIP, 3.63 ERA
MLB.com: 10-13, 193 IP, 137 K, 32 BB, 1.23 WHIP, 3.96 ERA
ZiPS: 26 G, 167.2 IP, 115 K, 27 BB, 4.03 ERA
Steamer: 9-10, 27 G, 163 IP, 113 K, 27 BB, 4.17 ERA
ESPN: 10 W, 122 K, 1.23 WHIP, 4.00 ERA

Average: 9.6 – 10.5, 27 G, 168 IP, 127 K, 31.2 BB, 1.20 WHIP, 3.89 ERA

(Original Post Here)

And his actual 2015 numbers:
14-13, 33 G, 31 GS, 194.2 IP, 1.24 WHIP, 4.16 ERA, 3.87 FIP

Colon pitched way more than the computers thought he would, which I’m wondering if that is the computers trying to factor Gee into the rotation with Niese, Wheeler, Harvey, Syndergaard, deGrom and Matz. He’s the only pitcher outside of Parnell who performed worse than his projections, yet, as a bit of a preview to next season, the computers think Colon is going to do better in 2016 than his 2015 projections let alone his actual 2015 numbers.

Jacob deGrom ended 2014 with the rookie of the year award. At the start of 2014, he was behind Montero on the depth chart, but then blossomed overnight. Would he repeat in 2015? Here is what the computers thought before the season:

PECOTA (BP): 8-9, 25 G, 147.2 IP, 126 K, 42 BB, 1.19 WHIP, 3.47 ERA
PECOTA (MLB.com): 9-8, 144 IP, 121 K, 41 BB, 1.20 WHIP, 3.56 ERA
MLB.com: 12-8, 180 IP, 161 K, 49 BB, 1.17 WHIP, 3.15 ERA
ZiPS: 29 G, 174.1 IP, 159 K, 53 BB, 3.30 ERA
Steamer: 10-10, 29 G, 163.0 IP, 145 K, 52 BB, 3.92 ERA
ESPN: 12-9, 177 K, 1.18 WHIP, 3.25 ERA

Average: 10.2 – 8.8, 27.6 G, 161.2 IP, 1.19 WHIP, 3.44 ERA

(Original post here)

And his actual 2015 numbers:
14-8, 30 G, 191 IP, 0.98 WHIP, 2.54 ERA.

The computers thought deGrom’s 2.69 ERA was not repeatable in 2015. The lowest ERA he was projected to get was 3.15 MLB.com, which generally projects in the player’s favor. Instead deGrom went out and put up a 2.54 ERA, thus cementing the Mets rotation as a staff of aces. The computers wanted him to have a sophomore slump, but that was not in the cards.

Our closer! At the start of the year, the question was “would Mejia be the closer the entire season?”. There was the thought that the Mets might do a closer by committee between Familia and Mejia. The most likely thought before the season was to start off with Mejia and then see how it goes. Familia was the one who was projected to have better stuff but Mejia had success the previous year. Then Mejia failed three drug tests and Familia put together one of the best seasons a Met closer has ever had. How much of this did the computers see at the start of last year? First, here are the 2015 projections:

PECOTA (BP): 3-3, 1 SV, 24 G, 58.2 IP, 55 K, 27 BB, 1.34 WHIP, 4.01 ERA
PECOTA (MLB.com): 3-3, 7 SV, 66 IP, 55 K, 28 BB, 1.35 WHIP, 4.05 ERA
MLB.com: 3-2, 0 SV, 72 IP, 68 K, 27 BB, 1.24 WHIP, 3.25 ERA
ZiPS: 65 G, 65.1 IP, 63 K, 28 BB, 3.17 ERA
Steamer: 3-3, 6 SV, 55 G, 55 IP, 56 K, 23 BB, 3.55 ERA
ESPN: 5-3, 0 SV, 71 K, 1.33 WHIP, 3.00 ERA

Average: 3.4-2.8, 2.8 SV, 48 IP, 63.1 IP, 61.3 K, 26.6 BB, 1.32 WHIP, 3.51 ERA

(Originally Posted Here)

And his actual 2015 stats:
2-2, 43 Saves, 76 G, 78 IP, 1.00 WHIP, 1.85 ERA.

And of course the answer is probably what you thought: no, the computers had no idea about Familia. The saves are understandable, they were operating under the idea of Mejia. The work load seemed a little light, but I guess the computers didn’t account for how Collins uses a bullpen. Ultimately though, Familia just performed a lot better than anyone thought. His WHIP was ridiculous which led to a ridiculous ERA (and a more down to Earth FIP of 2.76).

You know what feels like ages ago? Dillon Gee on the Mets. His story with the Mets is essentially the accelerated Jon Niese story. Gee was moved the bullpen before the season even started, which caused some controversy. He was then put back into the rotation thanks to Wheeler needing surgery. Gee fell to injury himself and when he came back, his pitching just wasn’t that good. The kids on the farm came up and Gee went out west for the resto of the season. He now finds himself with the World Champion Kansas City Royals. Could the projections in 2015 predict his struggles? Let’s take a look at his projections:

PECOTA (BP): 7-8, 22 G, 132.2 IP, 107 K, 39 BB, 1.22 WHIP, 3.90 ERA
MLB.com: 6-6, 110 IP, 77 K, 34 BB, 1.26 WHIP, 3.93 ERA
PECOTA (MLB.com): 7-7, 121 IP, 98 K, 36 BB, 1.21 WHIP, 3.79 ERA
ZiPS: 23 G, 142 IP, 107 K, 40 BB, 4.18 ERA
Steamer: 5-7, 17 G, 96 IP, 70 K, 26 BB, 4.35 ERA
ESPN: 6-5, 73 K, 1.22 WHIP, 3.86 ERA

Average: 6.2-6.6, 20.6 G, 120.1 IP, 88.6 K, 35 BB, 1.23 WHIP, 4.00 ERA

(Originally posted here)

His actual numbers:
0-3, 39.2 innings, 1.66 WHIP, 5.90 ERA

The projections pretty much missed the mark on all cylinders with Gee. They way overestimated his playing time, mainly because they underestimated the playing time of the other young members of the staff and then there was his WHIP. Gee just couldn’t get it together, especially once he came back from injury. This isn’t really a failure on the projection software but the unfortunate reality of Gee’s life last season.

Last year I looked at every single Matt Harvey projection that I could find probably like every other Mets fan. Would we get old Matt Harvey back? It felt like the entire season was up to how Harvey would perform, which isn’t too far from the truth. Let’s take a look at those 2015 Harvey projections and compare it to his performance:

PECOTA (BP): 2-2, 7 G, 39 IP, 40 K, 11 BB, 1.09 WHIP, 2.85 ERA
PECOTA (MLB.com): 10-8, 151.1 IP, 152 K, 43 BB, 1.09 WHIP, 2.91 ERA
MLB.com: 10-7, 170.0 IP, 191 K, 37 BB, 1.08 WHIP, 3.07 ERA
ZiPS: 25 G, 153 IP, 140 K, 49 BB, 3.12 ERA
Steamer: 9-7, 23 G, 144 IP, 158 K, 39 BB, 3.00 ERA
ESPN: 11-6, 172 K, 1.10 WHIP, 3.04 ERA

Average: 8.4-6, 131.3 IP, 142 K, 35.8 BB, 1.09 WHIP, 2.998 ERA

(Original posted here)

And in 2015 he put up these numbers:
13-8, 29 G, 189.1 IP, 1.02 WHIP, 2.71 ERA

Harvey was great! His pitching numbers are higher than a lot of people thought because pitching limits were sort of enforced but sort of not. His ERA was better than any individual projection. His WHIP was accurately predicted by just about everyone which is actually amazing. As a quick preview to a later article, the projections are expecting much of the same for 2016.

Jon Niese started the season off as a starter and found his way in an out of the rotation around Steven Matz. Before being traded, the post-season previewed one possible universe for him in 2016 where he would have been a lefty specialist out of the pen. Ultimately the Mets traded him at the end of the season. The long-time Met had been part of the rotation for a while with fairly consistent numbers. The question is, did the computers accurately predict Niese in 2015? Let’s see what the computers said last year:

PECOTA (BP): 8-10, 26 G, 155 IP, 128 K, 40 BB, 1.23 WHIP, 3.85 ERA
PECOTA (MLB.com): 10-11, 171.0 IP, 139 K, 43 BB, 1.23 WHIP, 3.89 ERA
MLB.com: 10-12, 183.0 IP, 132 K, 48 BB, 1.29 WHIP, 3.69 ERA
ZiPS: 28 G, 171.2 IP, 132 K, 46 BB, 3.77 ERA
Steamer: 9-11, 29 G, 173.0 IP, 125 K, 47 BB, 4.07 ERA
ESPN: 11-9, 140 K, 1.23 WHIP, 3.72 ERA

Average: 9.6-10.6, 27.7 G, 170.2 IP, 132.7 K, 44.8 BB, 1.25 WHIP, 3.83 ERA

(Originally posted here)

And here are is 2015 numbers:
9-10, 33 G, 29 GS, 176.2 IP, 1.40 WHIP, 4.13 ERA

The computers pretty much nailed Niese’s win-loss record for last year, and the amount of work he put it in. They thought he would have a better WHIP and ERA. Niese allowed a lot of hits and walks last year. There were times that I was surprised his his earned runs were so low (this is partially why his FIP is higher than his ERA).

This was a season was a missed opportunity for Parnell. He was coming off of Tommy John and into a bullpen over the summer that just found out it would be without Mejia for the whole season. It’s difficult to project how a pitcher will be able to perform after Tommy John, so lets see how Parnell did:

PECOTA (BP): 2-1, 15 SV, 40 G, 39.2 IP, 37 K, 11 BB, 1.14 WHIP, 2.97 ERA
PECOTA (MLB.com): 2-2, 3 SV, 40 IP, 38 K, 12 BB, 1.15 WHIP, 3.15 ERA
MLB.com: 2-2, 14 SV, 42 IP, 34 K, 14 BB, 1.29 WHIP, 3.49 ERA
ZiPS: 47 G, 44.2 IP, 37 K, 13 BB, 3.43 ERA
Steamer: 3-3, 10 SV, 65 G, 65 IP, 60 K, 21 BB, 3.37 ERA
ESPN: 2-3, 27 SV, 46 K, 1.10 WHIP, 2.52 ERA

Average: 2.2-2.2, 13.8 SV, 50.7 G, 46 IP, 42 K, 14.2 BB, 1.17 WHIP, 3.16 ERA

(Originally posted here)

2015 Stats:
2-4, 30 G, 24 IP, 4.9 K/9, 1.96 WHIP, 6.38 ERA

Look, no one could really predict how much Parnell was going to struggle. Based on when he came back, he could have potentially hit the games pitched/innings projected but his problems on the mound forced the Mets to go find someone else to take over the setup role.

Noah was sensational for the Mets this year and for a period of time, he was in the running for the Rookie of the Year. He really stepped it up in the playoffs putting up great numbers, lighting up the radar gun and showing some good attitude on the mound. Projections at the start of the season were struggling to figure out how much playing time he was going to get, let’s see if they were right:

PECOTA (BP): 7-8, 25 G, 120.1 IP, 118 K, 36 BB, 1.20 WHIP, 3.50 ERA
PECOTA (MLB.com): 2-2, 26.2 IP, 26 K, 8 BB, 1.20 WHIP, 3.78 ERA
MLB.com: 4-3, 50 IP, 46 K, 16 BB, 1.28 WHIP, 3.42 ERA
ZiPS: 26 G, 134.1 IP, 142 K, 38 BB, 3.48 ERA
Steamer: 2-2, 7 G, 38.0 IP, 38 K, 12 BB, 3.61 ERA
ESPN: 1-2, 33 K, 1.47 WHIP, 4.80 ERA

Average: 3.2-3.4, 19.3 G, 73.2 IP, 67 K, 27.2 BB, 1.29 WHIP, 3.77 ERA

(Originally posted here)

And here are is his 2015 numbers:
9-7, 24 G, 150 IP, 166 K, 1.05 WHIP, 3.24 ERA

Noah did great, he did way better than the projections thought he would, even the ones that were close to the amount he would actually pitch (PECOTA BP, ZiPS). He ended up with an ERA of about half a run less than projected and with a WHIP way better than predicted. It was a great first year for Syndergaard, but we already knew that.

The Mets signed Michael Cuddyer at the start of the 2015 off-season. On the day a chunk of the baseball world thought Cuddyer was going to take the qualifying offer from the Rockies, Cuddyer signed a two year deal with the Mets. He struggled with the Mets, especially towards the end of the season and in the post-season and retired leaving quite a bit of money on the table at the end of the season, which ended up being one of the major stories of the 2016 season (it is difficult seeing the Mets getting both de Aza and Cespedes with Cuddyer still on the team). Here are what the projections thought Cuddyer would do:

PECOTA (BP): 265 PA, 29 R, 14 2B, 1 3B, 7 HR, 31 RBI, .260 BA, .320 OBP
PECOTA (MLB.com): 465 AB, 57 R, 14 HR, 60 RBI, .258 BA, .319 OBP
MLB.com: 412 AB, 60 R, 14 HR, 60 RBI, .279 BA, .340 OBP
ZiPS: 372 PA, 42 R, 21 2B, 2 3B, 12 HR, 44 RBI, .271 BA, .325 OBP
Steamer: 597 PA, 543 AB, 63 R, 28 2B, 2 3B, 18 HR, 69 RBI, .254 BA, .313. OBP
ESPN: 66 R, 19 HR, 67 RBI, .265 BA

Average: 411 PA, 473 AB, 52.8 R, 21 2B, 1.7 3B, 14 HR, 55.2 RBI, .265 BA, .323 OBP

(Originally posted here)

And here are is 2015 numbers:
408 PA, 44 R, 18 2B, 10 HR, 41 RBI, .259 BA, .309 OBP

This was another player the computers had a hard time pinpointing how much playing time he would see. Also the models were all over the place with ESPN saying 19 homers and BP PECOTA saying 7. All in all, his numbers were slightly below what was projected. The gap feels smaller because it feels like Cuddyer was terrible for the Mets, but these numbers say otherwise. He wasn’t good for them on the field, but he was terrible either.

He had a tremendous career and hope he has a great retirement!